Validity of methods for analyzing urinary steroid data to detect ovulation in athletes

Area of Science:

• Reproductive endocrinology
• Sports medicine
• Biomarker analysis

Background:

• Subclinical hormonal abnormalities are common in female athletes, potentially affecting menstrual cycle regularity.
• Accurate ovulation detection is crucial for fertility and reproductive health monitoring.

Purpose of the Study:

• To evaluate the accuracy of ovulation detection algorithms in menstrual cycles with subclinical hormonal abnormalities.
• To determine if algorithm performance differs between exercising women and sedentary controls.

Main Methods:

• Compared the validity of five ovulation detection algorithms.
• Analyzed daily urine samples for estrone-3-glucuronide (E1G), pregnanediol-3-glucuronide (PDG), and luteinizing hormone (LH).
• Assessed algorithm sensitivity, specificity, and deviation from reference ovulation day.

Main Results:

• Most algorithms showed >80% sensitivity, with two exceptions.
• Specificities were generally below 70%, except for one algorithm.
• Baird's E1G/PDG ratio algorithm demonstrated the highest accuracy in ovulation day estimation.
• No significant differences in algorithm sensitivity or ovulation day deviation were found between exercising women and controls.

Conclusions:

• Ovulation detection algorithms exhibit similar validity in both exercising women and controls.
• These algorithms can be reliably applied to menstrual cycles with subclinical hormonal abnormalities.
• The findings support the use of these algorithms for monitoring ovulation in female athletes.